-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
      name:  <unnamed>
       log:  C:\Users\knm4642\Dropbox\Nha Meo Heo Beo\_JMP\WORKING FOLDER\Table VI_Effect by pre-transition researcher-pool quality.log
  log type:  text
 opened on:  18 Feb 2026, 23:46:23

. 
. //PREPARE DATA
. use "CEO transition event sample.dta", clear

. * construct additional variables
. gen f1allpat_highq = 0

. forval i = 6/10 {
  2.         replace f1allpat_highq = f1allpat_highq + f1allpat_citff_dec`i'
  3. }
(3,218 real changes made)
(3,772 real changes made)
(4,526 real changes made)
(5,383 real changes made)
(6,253 real changes made)

. gen ash_f1allpat_highq = asinh(f1allpat_highq)

. gen ash_efficiency = asinh(f1allpat/xrd)
(11,104 missing values generated)

. replace ash_efficiency = asinh(f1allpat) if xrd == 0
(11,104 real changes made)

. * set globals
. global firm = "firmage firmage2 ash_xrd"

. global ceo = "gender age age2 yrinco i.education"

. global eventsample = "!bothnonUS"

. global cluster = "mainethcode"

. 
. 
. //PREPARE TABLE
. eststo clear

. eststo col1: /// all patents
>         reghdfe ash_f1allpat c.trust_sd##c.resqual ${firm} ${ceo} if ${eventsample}, ///
>         a(eventid year) cluster(${cluster}) keepsin
WARNING: Singleton observations not dropped; statistical significance is biased (link)
(MWFE estimator converged in 7 iterations)
note: resqual is probably collinear with the fixed effects (all partialled-out values are close to zero; tol = 1.0e-09)

HDFE Linear regression                            Number of obs   =     19,506
Absorbing 2 HDFE groups                           F(  12,     34) =      51.62
Statistics robust to heteroskedasticity           Prob > F        =     0.0000
                                                  R-squared       =     0.9094
                                                  Adj R-squared   =     0.8973
                                                  Within R-sq.    =     0.0103
Number of clusters (mainethcode) =         35     Root MSE        =     0.5515

                                   (Std. err. adjusted for 35 clusters in mainethcode)
--------------------------------------------------------------------------------------
                     |               Robust
        ash_f1allpat | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
---------------------+----------------------------------------------------------------
            trust_sd |    .063442   .0169808     3.74   0.001      .028933    .0979511
             resqual |          0  (omitted)
                     |
c.trust_sd#c.resqual |   .0454573   .0184106     2.47   0.019     .0080424    .0828721
                     |
             firmage |  -.0103065   .0015472    -6.66   0.000    -.0134507   -.0071622
            firmage2 |   .0000717   .0000894     0.80   0.428    -.0001099    .0002534
             ash_xrd |   .1044394   .0128263     8.14   0.000     .0783733    .1305055
              gender |  -.0344516   .0319702    -1.08   0.289    -.0994229    .0305198
                 age |  -.0033345   .0059964    -0.56   0.582    -.0155206    .0088516
                age2 |   .0000109    .000053     0.20   0.839    -.0000969    .0001187
              yrinco |   .0025341   .0005911     4.29   0.000     .0013328    .0037354
                     |
           education |
                  2  |   .0746682   .0395362     1.89   0.068     -.005679    .1550154
                  3  |   .0815163   .0370148     2.20   0.035     .0062933    .1567393
                  4  |   .1211957   .0273214     4.44   0.000      .065672    .1767195
                     |
               _cons |   .8456931   .1636849     5.17   0.000     .5130452    1.178341
--------------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     eventid |      2278           0        2278     |
        year |        13           1          12     |
-----------------------------------------------------+

.                 estadd local FE         "X"

added macro:
                 e(FE) : "X"

.                 estadd local controls   "X"

added macro:
           e(controls) : "X"

.                 distinct eventid if e(sample)

         |        Observations
         |      total   distinct
---------+----------------------
 eventid |      19506       2278

.                 eststo col1, add(noevents r(ndistinct))
(e(noevents) = 2278 added)

. eststo col2: /// top-quality patents
>         reghdfe ash_f1allpat_highq c.trust_sd##c.resqual ${firm} ${ceo} if ${eventsample}, ///
>         a(eventid year) cluster(${cluster}) keepsin
WARNING: Singleton observations not dropped; statistical significance is biased (link)
(MWFE estimator converged in 7 iterations)
note: resqual is probably collinear with the fixed effects (all partialled-out values are close to zero; tol = 1.0e-09)

HDFE Linear regression                            Number of obs   =     19,506
Absorbing 2 HDFE groups                           F(  12,     34) =      34.91
Statistics robust to heteroskedasticity           Prob > F        =     0.0000
                                                  R-squared       =     0.9045
                                                  Adj R-squared   =     0.8917
                                                  Within R-sq.    =     0.0070
Number of clusters (mainethcode) =         35     Root MSE        =     0.5302

                                   (Std. err. adjusted for 35 clusters in mainethcode)
--------------------------------------------------------------------------------------
                     |               Robust
  ash_f1allpat_highq | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
---------------------+----------------------------------------------------------------
            trust_sd |   .0549003    .014733     3.73   0.001     .0249593    .0848413
             resqual |          0  (omitted)
                     |
c.trust_sd#c.resqual |   .0435105   .0186112     2.34   0.025     .0056879    .0813331
                     |
             firmage |  -.0085823   .0018798    -4.57   0.000    -.0124025   -.0047621
            firmage2 |  -.0000315   .0000834    -0.38   0.708    -.0002011     .000138
             ash_xrd |   .0772056   .0120962     6.38   0.000     .0526232     .101788
              gender |  -.0222153    .027024    -0.82   0.417    -.0771345     .032704
                 age |  -.0006306   .0057401    -0.11   0.913    -.0122959    .0110346
                age2 |  -9.10e-06    .000054    -0.17   0.867    -.0001188    .0001006
              yrinco |   .0024515   .0006589     3.72   0.001     .0011124    .0037906
                     |
           education |
                  2  |   .0888173   .0467974     1.90   0.066    -.0062864    .1839211
                  3  |   .0856079   .0443924     1.93   0.062    -.0046083    .1758242
                  4  |   .1184837   .0315351     3.76   0.001     .0543968    .1825707
                     |
               _cons |   .7533603    .157098     4.80   0.000     .4340987    1.072622
--------------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     eventid |      2278           0        2278     |
        year |        13           1          12     |
-----------------------------------------------------+

.                 estadd local FE         "X"

added macro:
                 e(FE) : "X"

.                 estadd local controls   "X"

added macro:
           e(controls) : "X"

.                 distinct eventid if e(sample)

         |        Observations
         |      total   distinct
---------+----------------------
 eventid |      19506       2278

.                 eststo col2, add(noevents r(ndistinct))
(e(noevents) = 2278 added)

. eststo col3: /// R&D efficiency
>         reghdfe ash_efficiency c.trust_sd##c.resqual ${firm} ${ceo} if ${eventsample}, ///
>         a(eventid year) cluster(${cluster}) keepsin
WARNING: Singleton observations not dropped; statistical significance is biased (link)
(MWFE estimator converged in 7 iterations)
note: resqual is probably collinear with the fixed effects (all partialled-out values are close to zero; tol = 1.0e-09)

HDFE Linear regression                            Number of obs   =     19,506
Absorbing 2 HDFE groups                           F(  12,     34) =      71.65
Statistics robust to heteroskedasticity           Prob > F        =     0.0000
                                                  R-squared       =     0.7274
                                                  Adj R-squared   =     0.6910
                                                  Within R-sq.    =     0.0964
Number of clusters (mainethcode) =         35     Root MSE        =     0.3428

                                   (Std. err. adjusted for 35 clusters in mainethcode)
--------------------------------------------------------------------------------------
                     |               Robust
      ash_efficiency | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
---------------------+----------------------------------------------------------------
            trust_sd |   .0375853   .0126993     2.96   0.006     .0117773    .0633934
             resqual |          0  (omitted)
                     |
c.trust_sd#c.resqual |   .0278626   .0109606     2.54   0.016      .005588    .0501371
                     |
             firmage |  -.0044185   .0021346    -2.07   0.046    -.0087566   -.0000804
            firmage2 |  -9.92e-07   .0000422    -0.02   0.981    -.0000867    .0000847
             ash_xrd |  -.2455948   .0180376   -13.62   0.000    -.2822515   -.2089381
              gender |  -.0264704    .022551    -1.17   0.249    -.0722995    .0193586
                 age |  -.0075813   .0034334    -2.21   0.034    -.0145588   -.0006038
                age2 |   .0000588   .0000313     1.88   0.069    -4.80e-06    .0001225
              yrinco |    .001482   .0003837     3.86   0.000     .0007022    .0022618
                     |
           education |
                  2  |   .0040792   .0242147     0.17   0.867    -.0451311    .0532894
                  3  |   .0012514   .0221095     0.06   0.955    -.0436806    .0461834
                  4  |   .0315225   .0225055     1.40   0.170    -.0142142    .0772592
                     |
               _cons |   .9119887   .1375407     6.63   0.000     .6324724    1.191505
--------------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     eventid |      2278           0        2278     |
        year |        13           1          12     |
-----------------------------------------------------+

.                 estadd local FE         "X"

added macro:
                 e(FE) : "X"

.                 estadd local controls   "X"

added macro:
           e(controls) : "X"

.                 distinct eventid if e(sample)

         |        Observations
         |      total   distinct
---------+----------------------
 eventid |      19506       2278

.                 eststo col3, add(noevents r(ndistinct))
(e(noevents) = 2278 added)

. eststo col4: /// all patents
>         reghdfe ash_f1allpat c.trust_sd#i.resqualpct ${firm} ${ceo} if ${eventsample}, ///
>         a(eventid year) cluster(${cluster}) keepsin
WARNING: Singleton observations not dropped; statistical significance is biased (link)
(MWFE estimator converged in 7 iterations)

HDFE Linear regression                            Number of obs   =     19,506
Absorbing 2 HDFE groups                           F(  15,     34) =      70.36
Statistics robust to heteroskedasticity           Prob > F        =     0.0000
                                                  R-squared       =     0.9094
                                                  Adj R-squared   =     0.8973
                                                  Within R-sq.    =     0.0100
Number of clusters (mainethcode) =         35     Root MSE        =     0.5516

                                    (Std. err. adjusted for 35 clusters in mainethcode)
---------------------------------------------------------------------------------------
                      |               Robust
         ash_f1allpat | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------------+----------------------------------------------------------------
resqualpct#c.trust_sd |
                   1  |   .0145817     .03891     0.37   0.710     -.064493    .0936565
                   2  |   .0319719   .0269241     1.19   0.243    -.0227445    .0866883
                   3  |   .0642823   .0502628     1.28   0.210     -.037864    .1664286
                   4  |   .0843191   .0294437     2.86   0.007     .0244823    .1441559
                   5  |   .1301109   .0394142     3.30   0.002     .0500117    .2102102
                      |
              firmage |  -.0102528   .0017618    -5.82   0.000    -.0138331   -.0066724
             firmage2 |    .000072   .0000885     0.81   0.422     -.000108    .0002519
              ash_xrd |   .1043864   .0129128     8.08   0.000     .0781444    .1306284
               gender |  -.0342527    .032352    -1.06   0.297    -.0999999    .0314944
                  age |  -.0032297   .0061262    -0.53   0.601    -.0156796    .0092203
                 age2 |   9.85e-06   .0000541     0.18   0.856       -.0001    .0001197
               yrinco |   .0025454   .0006232     4.08   0.000     .0012789    .0038119
                      |
            education |
                   2  |   .0754166   .0379082     1.99   0.055    -.0016221    .1524553
                   3  |   .0820533    .035376     2.32   0.027     .0101606    .1539459
                   4  |   .1222216   .0267604     4.57   0.000     .0678379    .1766053
                      |
                _cons |   .8275864   .1730941     4.78   0.000     .4758168    1.179356
---------------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     eventid |      2278           0        2278     |
        year |        13           1          12     |
-----------------------------------------------------+

.                 estadd local FE         "X"

added macro:
                 e(FE) : "X"

.                 estadd local controls   "X"

added macro:
           e(controls) : "X"

.                 distinct eventid if e(sample)

         |        Observations
         |      total   distinct
---------+----------------------
 eventid |      19506       2278

.                 eststo col4, add(noevents r(ndistinct))
(e(noevents) = 2278 added)

. eststo col5: /// future sales
>         reghdfe f2logsale c.trust_sd##c.resqual ${firm} ${ceo} if ${eventsample}, ///
>         a(eventid year) cluster(${cluster}) keepsin
WARNING: Singleton observations not dropped; statistical significance is biased (link)
(MWFE estimator converged in 7 iterations)
note: resqual is probably collinear with the fixed effects (all partialled-out values are close to zero; tol = 1.0e-09)

HDFE Linear regression                            Number of obs   =     17,873
Absorbing 2 HDFE groups                           F(  12,     33) =      42.34
Statistics robust to heteroskedasticity           Prob > F        =     0.0000
                                                  R-squared       =     0.9649
                                                  Adj R-squared   =     0.9598
                                                  Within R-sq.    =     0.0302
Number of clusters (mainethcode) =         34     Root MSE        =     0.4907

                                   (Std. err. adjusted for 34 clusters in mainethcode)
--------------------------------------------------------------------------------------
                     |               Robust
           f2logsale | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
---------------------+----------------------------------------------------------------
            trust_sd |  -.0361679   .0173091    -2.09   0.044    -.0713834   -.0009523
             resqual |          0  (omitted)
                     |
c.trust_sd#c.resqual |   .0450339   .0140494     3.21   0.003     .0164501    .0736177
                     |
             firmage |  -.0361756   .0056099    -6.45   0.000    -.0475891   -.0247621
            firmage2 |  -.0004769   .0000779    -6.12   0.000    -.0006353   -.0003185
             ash_xrd |   .1262517   .0266626     4.74   0.000     .0720062    .1804971
              gender |   .0164909   .0504179     0.33   0.746     -.086085    .1190669
                 age |   .0212883   .0097559     2.18   0.036     .0014399    .0411368
                age2 |  -.0001987   .0000851    -2.33   0.026    -.0003719   -.0000255
              yrinco |  -.0001272   .0009249    -0.14   0.891    -.0020089    .0017544
                     |
           education |
                  2  |  -.0013479   .0629768    -0.02   0.983    -.1294752    .1267793
                  3  |  -.0295569    .065941    -0.45   0.657    -.1637149    .1046011
                  4  |   .0792587   .0582496     1.36   0.183     -.039251    .1977684
                     |
               _cons |   6.666553   .2796239    23.84   0.000     6.097653    7.235452
--------------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     eventid |      2229           0        2229     |
        year |        13           1          12     |
-----------------------------------------------------+

.                 estadd local FE         "X"

added macro:
                 e(FE) : "X"

.                 estadd local controls   "X"

added macro:
           e(controls) : "X"

.                 distinct eventid if e(sample)

         |        Observations
         |      total   distinct
---------+----------------------
 eventid |      17873       2229

.                 eststo col5, add(noevents r(ndistinct)) 
(e(noevents) = 2229 added)

. eststo col6: /// future employment      
>         reghdfe f2logemp c.trust_sd##c.resqual ${firm} ${ceo} if ${eventsample}, ///
>         a(eventid year) cluster(${cluster}) keepsin
WARNING: Singleton observations not dropped; statistical significance is biased (link)
(MWFE estimator converged in 7 iterations)
note: resqual is probably collinear with the fixed effects (all partialled-out values are close to zero; tol = 1.0e-09)

HDFE Linear regression                            Number of obs   =     17,684
Absorbing 2 HDFE groups                           F(  12,     33) =      18.78
Statistics robust to heteroskedasticity           Prob > F        =     0.0000
                                                  R-squared       =     0.9759
                                                  Adj R-squared   =     0.9724
                                                  Within R-sq.    =     0.0318
Number of clusters (mainethcode) =         34     Root MSE        =     0.3756

                                   (Std. err. adjusted for 34 clusters in mainethcode)
--------------------------------------------------------------------------------------
                     |               Robust
            f2logemp | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
---------------------+----------------------------------------------------------------
            trust_sd |  -.0347802   .0125587    -2.77   0.009    -.0603311   -.0092292
             resqual |          0  (omitted)
                     |
c.trust_sd#c.resqual |   .0348139    .009573     3.64   0.001     .0153374    .0542904
                     |
             firmage |  -.0228132   .0053976    -4.23   0.000    -.0337947   -.0118317
            firmage2 |  -.0004193   .0000689    -6.08   0.000    -.0005596   -.0002791
             ash_xrd |   .0849338   .0241057     3.52   0.001     .0358903    .1339773
              gender |  -.0741431   .0378046    -1.96   0.058    -.1510571    .0027709
                 age |    .018934   .0058582     3.23   0.003     .0070154    .0308526
                age2 |  -.0001891   .0000533    -3.55   0.001    -.0002976   -.0000806
              yrinco |   .0009152   .0008314     1.10   0.279    -.0007762    .0026066
                     |
           education |
                  2  |  -.0064121   .0594644    -0.11   0.915    -.1273935    .1145692
                  3  |  -.0230374   .0556141    -0.41   0.681    -.1361851    .0901103
                  4  |     .05874   .0576281     1.02   0.315    -.0585052    .1759851
                     |
               _cons |    1.01403   .2559306     3.96   0.000     .4933351    1.534725
--------------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     eventid |      2216           0        2216     |
        year |        13           1          12     |
-----------------------------------------------------+

.                 estadd local FE         "X"

added macro:
                 e(FE) : "X"

.                 estadd local controls   "X"

added macro:
           e(controls) : "X"

.                 distinct eventid if e(sample)

         |        Observations
         |      total   distinct
---------+----------------------
 eventid |      17684       2216

.                 eststo col6, add(noevents r(ndistinct)) 
(e(noevents) = 2216 added)

. eststo col7: /// future TFP
>         reghdfe f2TFP_KL c.trust_sd##c.resqual ${firm} ${ceo} if ${eventsample}, ///
>         a(eventid year) cluster(${cluster}) keepsin
WARNING: Singleton observations not dropped; statistical significance is biased (link)
(MWFE estimator converged in 7 iterations)
note: resqual is probably collinear with the fixed effects (all partialled-out values are close to zero; tol = 1.0e-09)

HDFE Linear regression                            Number of obs   =     17,125
Absorbing 2 HDFE groups                           F(  12,     33) =       6.04
Statistics robust to heteroskedasticity           Prob > F        =     0.0000
                                                  R-squared       =     0.7697
                                                  Adj R-squared   =     0.7359
                                                  Within R-sq.    =     0.0025
Number of clusters (mainethcode) =         34     Root MSE        =     0.3721

                                   (Std. err. adjusted for 34 clusters in mainethcode)
--------------------------------------------------------------------------------------
                     |               Robust
            f2TFP_KL | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
---------------------+----------------------------------------------------------------
            trust_sd |   .0035028    .012688     0.28   0.784    -.0223113    .0293168
             resqual |          0  (omitted)
                     |
c.trust_sd#c.resqual |   .0131411   .0108608     1.21   0.235    -.0089553    .0352376
                     |
             firmage |   .0010344   .0080078     0.13   0.898    -.0152576    .0173263
            firmage2 |  -.0000726   .0000298    -2.44   0.020    -.0001332    -.000012
             ash_xrd |    .027102   .0122332     2.22   0.034     .0022134    .0519905
              gender |   .0091125   .0215863     0.42   0.676    -.0348052    .0530302
                 age |   .0091342   .0085782     1.06   0.295    -.0083182    .0265867
                age2 |  -.0000788   .0000754    -1.05   0.303    -.0002322    .0000746
              yrinco |  -.0008141   .0005982    -1.36   0.183    -.0020311    .0004028
                     |
           education |
                  2  |  -.0219801   .0166398    -1.32   0.196     -.055834    .0118737
                  3  |  -.0422935   .0217842    -1.94   0.061    -.0866138    .0020269
                  4  |  -.0069072   .0237113    -0.29   0.773    -.0551482    .0413338
                     |
               _cons |  -.1244094   .3294774    -0.38   0.708    -.7947362    .5459175
--------------------------------------------------------------------------------------

Absorbed degrees of freedom:
-----------------------------------------------------+
 Absorbed FE | Categories  - Redundant  = Num. Coefs |
-------------+---------------------------------------|
     eventid |      2168           0        2168     |
        year |        13           1          12     |
-----------------------------------------------------+

.                 estadd local FE         "X"

added macro:
                 e(FE) : "X"

.                 estadd local controls   "X"

added macro:
           e(controls) : "X"

.                 distinct eventid if e(sample)

         |        Observations
         |      total   distinct
---------+----------------------
 eventid |      17125       2168

.                 eststo col7, add(noevents r(ndistinct))
(e(noevents) = 2168 added)

. 
. lab var ash_f1allpat        "All patents"

. lab var ash_f1allpat_highq  "Top qual patents"

. lab var ash_efficiency      "R\&D efficiency"

. lab var f2logsale           "ln(sales)"

. lab var f2logemp            "ln(employment)"

. lab var f2TFP_KL            "TFP"

. global coeflist1 = "trust_sd c.trust_sd#c.resqual"

. global coeflist2 = "1.resqualpct#c.trust_sd 2.resqualpct#c.trust_sd 3.resqualpct#c.trust_sd 4.resqualpct#c.trust_sd 5.resqualpct#c.trust_sd"

. esttab /*using "Table VI_Effect by pre-transition researcher-pool quality.tex"*/, ///
>         cells(b(star fmt(3)) se(par fmt(3))) starlevels(* 0.1 ** 0.05 *** 0.01) ///
>         collabels(none) label varwidth(33) modelwidth(12) ///
>         mgroups("1-year" "2-year", pattern(1 0 0 0 1 0 0) ///
>                 prefix(\multicolumn{@span}{c}{) suffix(}) span) ///
>         keep(${coeflist1} ${coeflist2}) order(${coeflist1} ${coeflist2}) ///    
>         coeflab(trust_sd "CEO's trust" ///
>                         c.trust_sd#c.resqual "Trust $\times$ Researcher quality" ///
>                         1.resqualpct#c.trust_sd "Trust $\times$ Quality quintile 1" ///
>                         2.resqualpct#c.trust_sd "Trust $\times$ Quality quintile 2" ///
>                         3.resqualpct#c.trust_sd "Trust $\times$ Quality quintile 3" ///
>                         4.resqualpct#c.trust_sd "Trust $\times$ Quality quintile 4" ///
>                         5.resqualpct#c.trust_sd "Trust $\times$ Quality quintile 5") ///
>         stats(FE controls N noevents, fmt(%9.0fc %9.0fc) ///
>                 lab("Event \& Year FEs" "Baseline controls" "Observations" "Events"))

-------------------------------------------------------------------------------------------------------------------------------------------------
                                  \multicolumn{4}{c}{1-year}                                      \multicolumn{3}{c}{2-year}                     
                                           (1)             (2)             (3)             (4)             (5)             (6)             (7)   
                                   All patents    Top qual p~s    R\&D effic~y     All patents       ln(sales)    ln(employm~)             TFP   
-------------------------------------------------------------------------------------------------------------------------------------------------
CEO's trust                              0.063***        0.055***        0.038***                       -0.036**        -0.035***        0.004   
                                       (0.017)         (0.015)         (0.013)                         (0.017)         (0.013)         (0.013)   
Trust $\times$ Researcher quality        0.045**         0.044**         0.028**                         0.045***        0.035***        0.013   
                                       (0.018)         (0.019)         (0.011)                         (0.014)         (0.010)         (0.011)   
Trust $\times$ Quality quintile 1                                                        0.015                                                   
                                                                                       (0.039)                                                   
Trust $\times$ Quality quintile 2                                                        0.032                                                   
                                                                                       (0.027)                                                   
Trust $\times$ Quality quintile 3                                                        0.064                                                   
                                                                                       (0.050)                                                   
Trust $\times$ Quality quintile 4                                                        0.084***                                                
                                                                                       (0.029)                                                   
Trust $\times$ Quality quintile 5                                                        0.130***                                                
                                                                                       (0.039)                                                   
-------------------------------------------------------------------------------------------------------------------------------------------------
Event \& Year FEs                            X               X               X               X               X               X               X   
Baseline controls                            X               X               X               X               X               X               X   
Observations                            19,506          19,506          19,506          19,506          17,873          17,684          17,125   
Events                                   2,278           2,278           2,278           2,278           2,229           2,216           2,168   
-------------------------------------------------------------------------------------------------------------------------------------------------

. 
. cap log close
